Buck-boost transformers: small voltage adjustment field guide

What a buck-boost transformer is and why you reach for one

A buck-boost transformer is a small insulating transformer connected as an autotransformer to adjust the voltage a little. Boost raises it, buck lowers it, and the change is a fixed percentage of the supply. The job is not to convert the voltage to something new or to isolate the load. The job is to trim a supply that is close but not quite right for the equipment hanging off it.

Think of it as a correction, not a conversion. A transformer that changes 480 V to 208 V does real work on the whole load. A buck-boost moves 208 V to 230 V, or 240 V down to 208 V, by a small step, and that small step is all it has to handle. The companion guide on transformer types covers the full-conversion units and how the windings couple magnetically; this guide is the small-adjustment case, where the transformer is wired so the supply and the output share a winding.

You reach for one when the equipment nameplate and the measured supply do not line up, and the gap is small enough that a full distribution transformer would be overkill and overpriced. A motor that wants 230 V on a 208 V service is the textbook case. So is a long run that arrives low at the far end, or a piece of imported gear built for a voltage your service does not produce.

The problem it solves: a supply that is a little off

Most buck-boost work comes from one of three situations, and they all share the same shape. The supply is in the neighborhood of what the equipment wants, but off by enough to cause trouble.

The common one is a 208 V supply feeding equipment that wants 230 V or 240 V. A lot of three-phase services land at 208 V, and a lot of motors and packaged units are built for 230 V. The gear runs, but it runs unhappy. The second is low voltage at the end of a long run, where conductor resistance has eaten the supply down below what the load needs by the time it arrives, and upsizing the conductor is not practical after the fact. The third is the opposite, a supply that sits high and shortens the life of equipment that does not like it.

In each case the fix is a small trim. Boost the 208 up toward 230, boost the sagged far-end voltage back up, or buck the high supply down into range. The buck-boost moves the voltage the small amount needed and leaves the rest alone. That is the whole problem it is built for, and it is worth being honest that this is all it does.

An insulating transformer connected as an autotransformer

Here is the part that trips people up. A buck-boost transformer ships from the factory as a standard insulating transformer with separate primary and secondary windings, typically a low-voltage isolation unit. You do not use it that way. In the field you tie a primary lead to a secondary lead, which links the two windings electrically and turns the unit into an autotransformer.

An autotransformer has its primary and secondary connected, so they share part of a winding instead of being isolated. Current flows partly through the magnetic coupling and partly straight through the shared connection. That direct path is exactly why a buck-boost in this connection does not isolate the load. The output is referenced to the supply, not separated from it, which is a real difference from a full isolation transformer and matters for grounding.

So the same physical box is two different devices depending on how it is wired. Wired with the windings separate, it isolates and changes the full voltage at its rated low kVA. Wired as an autotransformer, it adjusts the voltage a small amount and carries a much larger load. The buck-boost is the second connection, and the nameplate kVA you see is the isolation rating, not the load it can support as an autotransformer.

Why a small kVA unit corrects a large load

This is the reason buck-boost transformers are cheap, and it is worth understanding rather than just trusting. In the autotransformer connection the transformer only processes the difference voltage, not the full load voltage. The bulk of the power flows straight through the shared connection and never gets transformed.

Put numbers on it. Boost 208 V to 230 V and the transformer is only working on the 22 V of difference, roughly a tenth of the voltage. So it only has to handle about a tenth of the load power. A unit with a 1 kVA isolation nameplate, connected to boost 208 V to 230 V, can support something on the order of 9 to 10 kVA of load, because nine tenths of that load power bypasses the windings. Manufacturer literature shows that same single unit jumping from a 1 kVA isolation rating to roughly 9.5 kVA in the boost connection.

The rule behind it is simple: the smaller the percent change, the larger the load a given unit handles. A 5 percent trim leans on the windings even less than a 10 percent trim. That ratio is the entire economic case for a buck-boost. You buy a small transformer and it does a large job, because it is only ever asked to handle the slice of the voltage you are changing.

How do you size a buck-boost transformer?

Size the transformer to the difference voltage it carries, not to the full load. The throughput the windings see is roughly the load current times the volts you are adding or subtracting, divided by 1000 for kVA. That number is small, which is the point, but the practical sizing is done off the manufacturer's selection table, not off a hand calculation.

Work an example. A 28 A load that you are boosting from 208 V to 230 V is adding about 22 V. The transformer throughput is about 28 times 22 divided by 1000, or roughly 0.62 kVA, so you would land on the next standard size up, commonly a 0.75 kVA unit. That small unit is serving a load of about 28 times 230 divided by 1000, near 6.4 kVA. A 0.75 kVA transformer supporting a 6.4 kVA load is normal for this connection, not a typo.

The reason to use the manufacturer table instead of trusting the bare calculation is that the table accounts for the exact winding voltages, the connection you pick, and the standard kVA steps the unit actually comes in. Manufacturers like Acme, Eaton, Hammond, Sola, and others publish tables that take your supply, your target, and your load current and hand back the model and the connection diagram. Use the table for the buy, and use the difference-voltage math as a sanity check that the table answer is in the right range.

Standard sizes and the 12/24 V windings

The workhorse single-phase buck-boost is built on a 120 by 240 V primary with a 12/24 V secondary, often written 120x240 primary, 12/24 secondary. That four-winding design is what lets a single model serve many supply and target combinations, because the windings can be put in series or parallel and tied to the line several different ways.

The two secondary coils give you a choice. Wire them in parallel and you get a 12 V winding to add or subtract. Wire them in series and you get 24 V. On a roughly 240 V line, 12 V is about a 5 percent change and 24 V is about a 10 percent change, so the same physical unit gives you two correction sizes depending on how you land the secondaries. A four-winding unit can be connected several distinct ways for different voltage and kVA combinations.

Larger corrections and other supply voltages use different winding voltages, but the principle holds: you pick the unit whose winding voltage matches the trim you need, then connect it for buck or boost. The manufacturer table maps your supply and target to the specific model and the specific connection, which is why the selection is a lookup, not a guess at the counter.

Buck versus boost: follow the connection diagram

Buck lowers the output, boost raises it, and which one you get depends entirely on how the secondary winding is phased against the supply. Connected so the secondary voltage adds to the line, you boost. Connected so it subtracts, you buck. Same transformer, same load, opposite result, decided by which leads you land.

This is the step where the work is won or lost, so the rule is blunt: follow the manufacturer connection diagram for the exact model and the exact buck or boost you want. The diagrams are specific about which winding leads go to line, which tie together in series or parallel, and which polarity makes the secondary additive or subtractive. A four-winding unit can be auto-connected several ways, and only the diagram for your supply and target is correct. Reading the diagram off a similar but different model is how a boost becomes a buck.

Get the polarity backward and the failure is loud and immediate. You meant to boost 208 toward 230 and instead you bucked it toward 186, and the equipment runs worse than it did on the raw supply. Verify the connection against the printed diagram before you energize, then verify the output with a meter, because the only way to be sure you boosted instead of bucked is to measure it.

Single-phase connections

A single-phase buck-boost uses one transformer wired between the supply and the load. The supply lands on the primary side, the secondary is phased to add or subtract, and the corrected voltage comes out to the equipment. For a single-phase two-wire or three-wire correction, the manufacturer diagram shows exactly which of the 120x240 primary leads and which 12/24 secondary leads to tie.

Most small jobs are single-phase: a 240 V circuit that needs trimming, a 120 V load that arrives low, a single piece of equipment with a voltage mismatch. The math and the connection are the simplest version of the device, and the same difference-voltage sizing applies. Pick the winding voltage for the percent you need, wire it per the diagram for buck or boost, and confirm the output under load before you call it done.

Three-phase connections: open delta and wye

Three-phase corrections use more than one single-phase buck-boost. The two common arrangements are open delta and wye, and the number of units differs. An open-delta connection uses two transformers and is the usual choice on a three-wire three-phase supply. A wye connection uses three transformers and should only be applied on a four-wire source with a grounded neutral.

The open-delta arrangement is popular because it corrects all three lines with only two units, which keeps the cost down on a balanced three-phase load. The wye arrangement uses one transformer per phase referenced to the neutral, which is why it needs that fourth wire to work correctly. Putting a wye connection on a three-wire delta source is a known mistake, and the manufacturers call it out specifically. If you are fuzzy on which service you have, the companion three-phase guide walks through telling wye from delta and reading line versus phase voltage before you commit to a connection.

As with single-phase, the manufacturer publishes separate selection tables and diagrams for open-delta and for wye. Confirm your service type first, then pull the matching table, because the unit count, the wiring, and the kVA per unit all change with the arrangement.

Typical percent of buck or boost

Buck-boost transformers are built for small corrections, commonly in the range of 5 to 20 percent, with 5, 10, and 15 percent being the everyday values. The 5 percent and 10 percent steps come straight out of the 12 V and 24 V secondaries on the standard 120x240 unit, which is why those two show up so often.

Pick the smallest percent that lands the equipment inside its nameplate range, because a smaller trim leans on the transformer less and lets a smaller unit do the job. Boosting 208 to about 230 is roughly a 10 percent move. Trimming 240 down to 208 is roughly a 13 percent move. The exact percent your unit delivers is fixed by its winding voltages and the connection, so confirm it against the manufacturer table rather than assuming a round number.

If the correction you need is larger than about 20 percent, a buck-boost is the wrong tool and you are into full transformer territory, where the unit has to handle the whole load power and the price and size climb accordingly.

It is a fixed change, not a voltage regulator

This is the misunderstanding that causes the most trouble, so be clear on it. A buck-boost transformer makes a fixed percentage change. It does not sense the supply and adjust. It does not hold the output steady. Whatever the supply does, the buck-boost adds or subtracts the same percentage and the output moves right along with it.

Boost a 208 V supply by 10 percent and you get about 229 V. If that supply sags to 200 V under load, the output drops to about 220 V, because the transformer is still just adding its fixed slice. If the supply swings high to 216 V, the output rides up to about 238 V. The buck-boost is a constant multiplier, not a thermostat for voltage.

So use a buck-boost when the supply is consistently off by a known, steady amount, and you want to shift it. When the supply swings around and you need the output held within a band regardless of what comes in, that is a voltage regulator or a tap-changing device, not a buck-boost. Pick the buck-boost for a fixed offset, and reach for a regulator when the problem is a moving supply. Confusing the two leaves you with equipment that is still out of range half the time.

Where buck-boost transformers actually get used

The applications all come back to a supply that is off by a small, steady amount. The most common is matching a 208 V service to 230 V or 240 V equipment, which covers a lot of motors, HVAC condensing units, heat pumps, and pumps that were built for 230 V and ended up on a 208 V building.

The next is boosting a circuit that arrives low at the end of a long run, where the voltage drop over the distance has pulled it under what the load needs and re-pulling larger conductor is not in the cards. Imported equipment built for a voltage the local service does not produce is another, where a small trim brings the supply into the gear's window. You also see buck-boost on lighting circuits that run high or low enough to shorten lamp life or dim output, and increasingly around EV charging and similar continuous loads where a long feeder leaves the voltage a little short at the equipment.

The thread through all of these is the same. The supply is close, the gap is steady, and the cheap fix is a small autotransformer trim rather than a new service or a full transformer. Where the supply swings instead of sitting steady, none of these is a buck-boost job.

The classic case: a 230 V motor on a 208 V supply

Running a 230 V motor on a 208 V supply is the application buck-boost transformers were practically made for, and it is worth knowing why the mismatch hurts. A motor fed below its rated voltage draws more current to produce the same torque, runs hotter, and over time that extra heat shortens the insulation life and can trip the overloads. The motor will start and run on 208 V, which is exactly why the problem gets ignored until the windings cook.

The fix is a boost connection that lifts the 208 V supply up toward 230 V, putting the motor back inside its nameplate range so it runs at its design current and temperature. Because the boost is only about 22 V, a small buck-boost handles a large motor, which is the whole appeal. A modest unit can correct a several-horsepower motor for a fraction of the cost of a full transformer.

One caution on motors specifically: they draw heavy inrush at start, several times running current for a moment. Size the buck-boost off the manufacturer table for the motor load, and confirm the boosted voltage holds up under running conditions, because a unit that looks fine at no load is not the test. The test is the motor running.

Grounding and overcurrent protection

Because the buck-boost is connected as an autotransformer, the output shares a connection with the supply and is not isolated from it. That changes how grounding is handled compared with an isolation transformer, and the autotransformer rules in the NEC apply rather than the separately-derived-system rules. Circuits supplied from autotransformers are addressed in the code, commonly cited around 210.9, and the autotransformer and grounding provisions live in Article 450, with grounding autotransformers covered near 450.5.

Overcurrent protection for the transformer follows the transformer article, Article 450, which sets how the primary and where required the secondary are protected based on the unit's rating. The exact percentages, the primary-only versus primary-and-secondary rules, and the conductor protection details depend on the rating and the connection, so size the overcurrent device per Article 450 for the unit you actually install.

The honest hedge: code section numbers and the specific grounding and overcurrent requirements shift between code cycles and are amended locally. Confirm the autotransformer grounding and overcurrent rules against the adopted NEC edition and any local amendments, and follow the manufacturer's installation instructions, before you finalize the protection and the bonding. Do not treat a buck-boost as a separately derived system, because in this connection it is not one.

Installation: measure the supply first

The first step is not picking the transformer. It is measuring the actual supply voltage, under load, at the point where the buck-boost will land. A service that is called 208 V might sit at 204 V or 213 V depending on the utility and the building load, and an open-circuit reading taken with nothing running can read several volts higher than what the equipment actually sees when the load is on. Size off the real number, measured under load, not the nominal label.

With the measured supply and the equipment's target voltage in hand, take the difference and the load current to the manufacturer selection table, pull the model and the connection diagram, and wire the unit exactly to that diagram for buck or boost. Land the primary leads and the secondary leads as drawn, get the series or parallel arrangement of the secondaries right, and confirm the polarity gives you the additive connection for boost or the subtractive one for buck.

Then energize and measure the output, again under load. Measuring first and measuring last bracket the job: the first reading sizes the unit and the last reading proves it did what you intended. Skipping the under-load supply reading is the most common way a buck-boost ends up the wrong size or correcting the wrong amount.

Verify the output under load

A buck-boost is only correct if the output lands where you need it with the load running, so the acceptance test is a meter reading under real load, not at idle. Voltage measured open-circuit at the output tells you the transformer changed something, but it does not tell you the equipment is in range once it pulls current and the supply sags.

Put the load on, or apply a known test load, and read the output voltage at the equipment terminals. Confirm two things: that the voltage moved in the direction you intended, boost up or buck down, and that the corrected value sits inside the equipment nameplate range with the load running. A clamp meter on the load conductor confirms the current you are correcting matches what you sized the unit for.

If the output is in the wrong direction, you wired a buck where you wanted a boost, and the fix is the connection, not the size. If the direction is right but the value is short, recheck the supply under load and the percent the unit actually delivers. Either way, the meter under load is the only thing that settles it.

How buck-boost installs go wrong

The failures cluster around a handful of repeat offenders. The most common is the wrong connection: you wanted boost and wired buck, so the equipment runs lower than it did on the raw supply. That comes from working off the wrong diagram or getting the secondary polarity backward, and it shows up the moment you meter the output.

Next is an undersized unit, usually from sizing off the nameplate isolation rating instead of the difference-voltage throughput, or from skipping the manufacturer table. A unit that is too small overheats and fails under the load it was supposed to correct. Then there is the regulator mistake, putting a fixed-change buck-boost on a supply that swings, so the output is in range sometimes and out of range the rest of the time.

The quieter failures are the measurement ones. Not reading the supply under load up front leaves you sizing off a nominal number that does not match reality. Getting the autotransformer grounding or overcurrent protection wrong creates a code and safety problem that energizes fine and bites later. And not verifying the output under load means you never actually confirmed the unit did its job. None of these is exotic, and all of them are caught by measuring first and measuring last.

Equipment voltage correction on data and process loads

Buck-boost shows up on data-center and process work for the same reason it shows up everywhere else: a piece of equipment wants a voltage the local service does not quite produce, and the gap is small and steady. Imported gear, legacy 230 V equipment on a 208 V hall, or a long feeder that arrives low at a distant rack row are the usual triggers.

The cautions are the same ones, just with less tolerance for getting them wrong. These loads are often continuous and sensitive, so verify the supply under load, size off the manufacturer table, and remember the buck-boost makes a fixed change. If the supply on that bus actually swings, a fixed trim will not hold a sensitive load in range, and the right tool is a regulator or a different power-conditioning device, not a buck-boost.

What to document

A buck-boost install is easy to misread later, because the same box can buck or boost and the only proof of which is the connection and the measured output. Write it down so the next person does not have to reverse-engineer it from the leads.

Capture the measured supply voltage under load, the equipment target voltage, the percent and direction of correction, the manufacturer model and the specific connection diagram used, whether it was wired buck or boost, the load current, the unit kVA, and the output voltage measured under load. Note the connection type for three-phase, open delta or wye, and the grounding and overcurrent arrangement. That record is what lets someone confirm months later that the correction is still right and was done to the diagram.

Common mistakes

• Wiring buck when you wanted boost, by working off the wrong diagram or reversing the secondary polarity.
• Sizing off the isolation nameplate instead of the difference-voltage throughput, ending up undersized.
• Treating a fixed-change buck-boost as a voltage regulator on a supply that swings.
• Not measuring the actual supply voltage under load before sizing the unit.
• Ignoring the autotransformer grounding and overcurrent rules and treating the output as isolated.
• Putting a wye connection on a three-wire source instead of an open delta.
• Not verifying the corrected output under real load after energizing.

Field checklist

Standards and references

The framework is the NEC, NFPA 70. Transformers are covered in Article 450, which includes autotransformer provisions and the overcurrent protection rules that apply to a buck-boost in its autotransformer connection. Circuits derived from or supplied by autotransformers are addressed elsewhere in the code, commonly cited around 210.9, and grounding autotransformers near 450.5. Because a buck-boost in this connection is an autotransformer and not a separately derived system, those are the rules that apply, not the isolation-transformer rules.

The buck-boost manufacturer is the second authority, and on the connection and sizing it is the controlling one. The selection table maps your supply, target, and load current to a specific model, kVA, and connection diagram, and that diagram dictates exactly which leads make buck versus boost. Acme, Eaton, Hammond, Sola, Schneider, and others publish these tables and diagrams, and the right answer is the one for your exact model. The equipment nameplate is the third reference, since it sets the target voltage and the tolerance you are correcting into.

Hedge the specifics. Code section numbers and the autotransformer grounding and overcurrent requirements change between cycles and are amended by jurisdiction, so confirm them against the adopted NEC edition and local amendments. Confirm the percent of correction and the kVA off the manufacturer table for the model in hand. And remember the two rules that matter most in the field: a small unit corrects a large load because it handles only the difference, you follow the manufacturer connection diagram for buck or boost, and the device makes a fixed change rather than regulating a swinging supply.

Units and terms

Buck-boost work carries a few terms that mean specific things, and mixing them up is how the wrong device gets ordered.

Buck means lower the voltage, boost means raise it, and both are a fixed percentage of the supply. An autotransformer is a transformer whose primary and secondary are electrically connected and share part of a winding, which is the connection a buck-boost uses in the field. The nameplate kVA is the unit's isolation rating, smaller than the load it supports as an autotransformer. Percent change is the trim as a fraction of the supply voltage, set by the winding voltages and the connection.

Buck / boost

Buck lowers the supply voltage, boost raises it, by a fixed percentage set by the connection

Autotransformer

A transformer with primary and secondary electrically tied, sharing a winding, so it does not isolate

Insulating transformer

The factory form of a buck-boost, with separate windings, before it is field-connected as an autotransformer

Nameplate kVA

The unit's isolation rating, much smaller than the load it supports in the buck-boost connection

Difference voltage

The volts added or subtracted, the only part the windings actually handle

Open delta / wye

Three-phase buck-boost arrangements using two units (open delta) or three units (wye, four-wire only)

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is the difference between a dry-type and liquid-filled transformer?
• What is the difference between wye and delta?
• How do you size a dry-type transformer?
• What is a low-voltage system?
• What is a stress cone?